DISSERTATION / DOCTORAL THESIS

Titel der Dissertation /Title of the Doctoral Thesis

„ Functional and mechanistic characterization of novel selective modulators of retinoid X receptors “

verfasst von / submitted by

Mag.rer.nat. Simone Latkolik

angestrebter akademischer Grad / in partial fulfilment of the requirements for the degree of

Doctor of Philosophy (PhD)

Wien, 2017/Vienna, 2017

Studienkennzahl lt. Studienblatt / degree programme code as it appears on the student record sheet:

A 794 685 490

Dissertationsgebiet lt. Studienblatt / field of study as it appears on the student record sheet:

Molekulare Biologie/Molecular Biology

Betreut von / Supervisor:

Univ. Prof. Dr. Verena Dirsch

“All my life through, the new sights of nature made me rejoice like a child.”

― Marie Curie

https://www.goodreads.com/author/show/126903.Marie_Curie

“Every great and deep difficulty bears in itself its own solution. It forces us to change our thinking in order to find it.”

-Niels Bohr

ABSTRACT

Retinoid X receptors (RXRs) are nuclear receptors displaying a variety of biological func-

tions. They regulate physiological and developmental processes by acting as signal inte-

grators to control the transcription of target genes.

Ligands of RXR are called rexinoids and they modulate nuclear receptor function either

as agonist or as antagonist. This modulation is due to the ligand-receptor complex for-

mation, the ability of RXRs to heterodimerize with other nuclear receptors or to form

homodimeric receptor complexes, as well as the ability of the formed ligand-receptor

complex to differentially recruit diverse co-activators and co-repressors to the respec-

tive target gene promoters. Some small molecule ligands exert anticancer activity and

have a potential for the treatment of metabolic diseases but their usage is limited due to

toxicity and unfavorable tissue and receptor subtype selectivity that cause various side

effects.

This work describes the identification of selective small-molecule ligands for RXR that

are structurally based on the partial RXR agonist and natural occurring rexinoid

Honokiol. Honokiol is a biphenylic neolignan initially isolated from the bark of Magnolia

officinalis. It has been shown that Honokiol promotes RXRα-, as well as PPARγ-

dependent luciferase gene expression in human embryonic kidney cells (HEK293). By

using receptor specific luciferase-based testing models, 5 asymmetric honokiol deriva-

tives were identified that selectively transactivate RXRα-dependent but not PPARγ-,

LXRα-, LXRβ-, and FXR-dependent luciferase gene expression in a dose-dependent man-

ner. The calculated EC50 values for RXRα activation indicate a high potency of action of

these derivatives and range from 170 to 850 nM in comparison to Honokiol (EC50: 1038

nM). Selected candidates were further functionally characterized and their impact on

macrophage cholesterol efflux in differentiated THP-1 cells as well as the impact on lipid

accumulation in liver cells (HepG2 cells) was investigated.

Additionally and in line with its partial agonism, a microarray assay for real-time co-

regulator-nuclear receptor interactions revealed that we have identified compounds

that recruit certain co regulators with a higher selectivity than full agonists (e.g. Bexaro-

tene, 9-cis retinoic acid) to the RXRα ligand-binding domain in vitro.

In summary, this work describes the functional and mechanistic characterization of

novel selective modulators of retinoid X receptors.

ZUSAMMENFASSUNG

Retinoid-X-Rezeptoren (RXRs) sind Kernrezeptoren, die eine Vielzahl von biologischen

Funktionen aufweisen. Sie regulieren physiologische Prozesse und spielen eine große

Rolle in der Entwicklung von Organsimen, indem sie als Signalintegratoren wirken um

die Transkription von Zielgenen zu kontrollieren.

Liganden von Retinoid-X-Rezeptoren werden als Rexinoide bezeichnet. Am Rezeptor

gebunden, agieren Sie entweder als Agonist oder Antagonist. Liganden Bindung führt

zur Bildung von Liganden-Rezeptor-Komplexen, die entweder Homodimere oder Hete-

rodimer sind, bestehend aus RXR und einem Partnerkernrezeptor. Die Rezeptoraktivie-

rung oder -hemmung hängt von verschiedenen Cofaktoren ab, die Liganden-abhängig

oder –unabhängig mit dem Rezeptor assoziiert sind und zu den jeweiligen Promotorre-

gionen der Zielgene rekrutieren werden.

RXR Liganden zeigen Aktivität gegen bestimmte Krebsarten und haben ein Potenzial zur

Behandlung von Stoffwechselerkrankungen. Die Möglichkeiten zur Anwendung dieser

RXR Liganden sind jedoch beschränkt durch einen hohen Toxizitätsgrad und, aufgrund

des breiten Wirkungsspektrums in verschiedenen Geweben, das Auftreten Nebenwir-

kungen. Diese Arbeit beschreibt die Identifizierung von neuen RXR Liganden mit selek-

tiveren Eigenschaften als volle Agonisten, die strukturell auf dem partiellen RXR-

Agonisten und dem natürlich vorkommenden Rexinoid Honokiol basieren.

Honokiol ist ein Biphenyl Neolignan, welches aus der Rinde von Magnolia officinalis iso-

liert wurde. Es wurde gezeigt, dass Honokiol die RXRα-, sowie PPARγ-abhängige Luci-

ferase-Genexpression in humanen embryonalen Nierenzellen (HEK293) erhöht. In ei-

nem zellbasierten Luciferase-Reportergensystem wurden asymmetrische Honokiolderi-

vate identifiziert, die selektiv die RXRα-abhängige, aber nicht PPARγ-, LXRα-, LXRβ- und

FXR-abhängige Luciferase-Genexpression in einer dosisabhängigen Weise transaktivie-

ren. Die berechneten EC50-Werte für die RXRα-Aktivierung zeigen eine hohe Potenz

(170- 850 nM) im Vergleich zu Honokiol (EC50: 1038 nM).

Die Wirkung dieser Derivate auf den Prozess des Cholesterin-Efflux in Makrophagen

sowie die Auswirkung auf die Lipidakkumulation in Leberzellen (HepG2-Zellen) wurde

genauer untersucht. Ein Mikroarray-Assay wurde durchgeführt, um die Interaktion von

Coregulator-Kernrezeptor-Wechselwirkungen zu identifizieren. Zusammenfassend be-

schreibt diese Arbeit die funktionale und mechanistische Charakterisierung von neuar-

tigen selektiven Modulatoren des Retinoid-X-Rezeptors.

CONTENTS

INTRODUCTION ................................................................................................................................. 1

1. NUCLEAR RECEPTORS .............................................................................................................. 1

1.1. THE NUCLEAR RECEPTOR SUPERFAMILY ................................................................................... 1

1.2. NUCLEAR RECEPTOR ARCHITECTURE ....................................................................................... 4

1.3. NUCLEAR RECEPTORS AS DRUG TARGETS ................................................................................. 6

1.4. THE RETINOID X RECEPTOR (RXR) ......................................................................................... 7

1.5. RETINOID X RECEPTOR LIGANDS (REXINOIDS) ......................................................................... 13

1.6. NUCLEAR RECEPTOR COREGULATORS .................................................................................... 20

2. TARGETING RETINOID X RECEPTORS IN HUMAN DISEASE ..................................... 22

2.1. THE ROLE OF NUCLEAR RECEPTORS IN MACROPHAGE BIOLOGY ................................................... 22

2.2. NUCLEAR RECEPTORS IN LIPID AND CHOLESTEROL METABOLISM ................................................. 23

2.3 ATHEROSCLEROSIS, LIPID- AND CHOLESTEROL METABOLISM ...................................................... 26

MATERIAL AND METHODS ......................................................................................................... 32

1. MATERIALS ............................................................................................................................... 32

1.1. PRODUCTS AND SUPPLIER INFORMATION ............................................................................... 32

1.2. CELL CULTURE MEDIA AND SUPPLEMENTS .............................................................................. 33

1.3. COMMERCIALLY AVAILABLE KITS .......................................................................................... 35

1.4. REAGENTS AND BUFFERS .................................................................................................... 35

1.5. PLASMID DNA ................................................................................................................ 38

1.6. PRIMER AND ANTIBODIES ................................................................................................... 39

1.7. TECHNICAL EQUIPMENT ..................................................................................................... 40

1.8. SCIENTIFIC SOFTWARE ....................................................................................................... 42

2. METHODS .................................................................................................................................. 43

2.1. PLASMID DNA PREPARATION ............................................................................................. 43

2.2. GENERAL CELL CULTURE CONDITIONS AND MAINTENANCE OF CELL LINES ...................................... 43

2.3. EXPERIMENTS IN HEK293 CELLS ......................................................................................... 45

2.4. RXRΑ CO-ACTIVATOR RECRUITMENT ASSAY ........................................................................... 49

2.5. REAL-TIME COREGULATOR- NUCLEAR RECEPTOR INTERACTION ................................................... 50

2.6. FUNCTIONAL STUDIES ........................................................................................................ 52

2.7. CELL VIABILITY ASSAY IN DIFFERENT CELL LINES (RESAZURIN ASSAY) ............................................ 59

2.8. RNA DETECTION BY QPCR ................................................................................................. 60

2.9. PROTEIN DETECTION BY WESTERN BLOTTING .......................................................................... 62

2.10. DATA ANALYSIS AND STATISTICS........................................................................................... 63

RESULTS AND DISCUSSION ......................................................................................................... 65

1. NUCLEAR RECEPTOR TRANSACTIVATION BY HONOKIOLDERIVATIVES ............ 65

1.1. SELECTIVE RXRΑ-DEPENDENT LUCIFERASE GENE TRANSACTIVATION BY HONOKIOL DERIVATIVES ....... 65

1.2. LUCIFERASE GENE TRANSACTIVATION OF HUMAN RXRΑ BY HONOKIOL DERIVATIVES 2284 AND 2817 IN

HEK293 CELLS .......................................................................................................................... 68

1.3. LUCIFERASE GENE TRANSACTIVATION OF RXRΑ BY 2284 AND 2817 IS ABOLISHED BY THE RXR-

ANTAGONIST HX531 .................................................................................................................. 69

1.4. NUCLEAR RECEPTOR LBD - GAL4-DBD/UAS-DEPENDENT LUCIFERASE GENE TRANSACTIVATION .... 70

1.5. SUMMARY AND DISCUSSION ............................................................................................... 73

2. CO-REGULATOR NUCLEAR RECEPTOR INTERACTION STUDIES WITH HUMAN

RXRΑ LBD ......................................................................................................................................... 75

2.1. RECRUITMENT OF PGC1Α TO HUMAN RXRΑ LBD BY 2284 IN VITRO ......................................... 75

2.2. COREGULATOR-NUCLEAR RECEPTOR INTERACTION PROFILING OF 2284 IN COMPARISON TO FULL RXR

AGONISTS ................................................................................................................................. 76

2.3. SUMMARY AND DISCUSSION ............................................................................................... 78

3. FUNCTIONAL STUDIES .......................................................................................................... 80

3.1. HONOKIOL DERIVATIVES 2284 AND 2817 PROMOTE CHOLESTEROL EFFLUX IN DIFFERENTIATED THP-1

MACROPHAGES .......................................................................................................................... 80

3.2. LIVER STEATOSIS MODEL IN HEPG2 CELLS .............................................................................. 84

3.3. ADIPOGENESIS IN 3T3-L1 CELLS .......................................................................................... 89

3.4. INFLUENCE OF HONOKIOL DERIVATIVES ON THE CELL VIABILITY IN DIFFERENT CELL LINES .................. 90

3.5. SUMMARY AND DISCUSSION ............................................................................................... 91

SUMMARY AND CONCLUSION .................................................................................................... 96

REFERENCES .................................................................................................................................. 101

APPENDIX ....................................................................................................................................... 117

ABBREVIATIONS ....................................................................................................................... 129

ACKNOWLEDGEMENTS/DANKSAGUNG ......................................................................................... 133

CURRICULUM VITAE .................................................................................................................. 135

Introduction __________________________________________________________________________________________________________________________

1

Introduction

1. Nuclear receptors Nuclear receptors are a super-family of highly conserved transcription factors that acti-

vate receptor-specific and tissue-specific sets of target genes. They are also called hor-

mone receptors because they were historically discovered in endocrinology and often

bind small lipophilic ligands like hormones. Modulation of nuclear receptor action has

been described in almost every aspect of development, cell-differentiation, metabolism,

cell death and even cancer and other human diseases.1

Because they regulate a diverse set of biological function with key roles in metabolic and

hormonal homeostasis virtually all nuclear receptors with identified ligands are well

characterized targets for drug development to treat various diseases including obesity,

diabetes, atherosclerosis, inflammation and endocrine disorders.2

1.1. The nuclear receptor superfamily In human 48 genes encode for nuclear receptors and the superfamily can be sub-divided

into 7 subfamilies including the receptors for lipophilic vitamins, cholesterol metabo-

lites, thyroid hormones, and steroid hormones.3,4

A large number of nuclear receptors are classified as orphan receptors because their

ligands are not discovered yet or are not characterized well.5-7

However, nuclear receptors with known, well-described ligands can be categorized as

follows: Classic steroid receptors are nuclear receptors for steroid hormones and func-

tion as homodimers. The androgen receptor (AR), estrogen receptor (ER), glucocorti-

coid receptor (GR), the mineralocorticoid receptor (MR) and the progesterone receptor

(PR) all belong to this category.8

Non-steroidal receptors function as hetero- or as homodimers to regulate downstream

gene expression. The central hetero-or homo- dimerization partner in this category of

nuclear receptors is the retinoid X receptor (RXR). The classic RXR heterodimer recep-

tors play important roles in metabolic homeostasis, development and immunity.9 An-

other category of nuclear receptors is the xenobiotic receptors that mainly has a role in

the protection of toxic substances.10,11 Table 1 summarizes the main categories.

Introduction __________________________________________________________________________________________________________________________

2

Category

Name

Subtypes/

Isoforms

Natural Ligand

Therapeutic

Relevance

Example of

therapeutic

Ligands

Classic RXR

Heterodimer Receptors

Retinoid X Receptor RXRα

RXRβ

RXRγ

All-trans retinoic

acid

Subcutaneous T-cell

lymphoma (Skin

Cancer)

Bexarotene

LG1069

(Targretin)

I. Permissive

Nuclear Receptors

Retinoic Acid Recep-

tor

RARα

RARβ

RARγ

Retinoic acid Acne Isotretinoin

(Accutane)

Liver X

Receptor

LXRα

LXRβ

24,25-Epoxycholes- terol, 24-Hydroxy- cholesterol

Atherosclerosis

(considered), Role in

Lipid and Cholesterol

synthesis

-

Peroxisome prolifer-

ator-activated

Receptors

PPARα

PPARβ/δ

PPARγ

Fatty acids,

Eicosanoids

Dyslipidemia

(PPARα), Diabetes

and Insulin sensitiza-

tion (PPARγ)

Fenofibrate

(Tricor;

PPARα),

Thiazolidenedi-

ones(Avandia,

Actos;PPARγ)

Farnesoid

Receptor

FXR Chenodeoxy-cholic acid

Role in cholesterol

maintenance, Choles-

tasis, Protects

hepatocytes from bile

toxicity

Obechitolic acid

(Ocaliva)

II. Non Permissive

Nuclear receptors

Vitamin D

Receptor

VDR

Vitamin D,

Bile acids

Hypocalcemia, Osteo-

porosis, Renal failure

Calcitriol

(Rocaltrol)

Thyroid hormone

Receptor

TRα

TRβ

Thyroid hor-

mone

Thyroid deficiency Levothyroixine

(Synthroid)

Classic Steroid Receptors Estrogen

Receptor

ERα

ERβ

Estrogens,

Estradiol

Breast Cancer, Osteo-

porosis prevention,

Menopausal symp-

toms

Tamoxifen,

Raloxifene

(Evista), Gen-

estein, Diethyl-

stilbestrol,

Equine estro-

genes

(Premarin)

Glucocorticoid Re-

ceptor

GR Glucocorticoids,

Cortisol

Asthma, Arthritis,

Rhinitis, Cancer,

Immune suppressant

Prednisone,

Dexamethasone

Mineralocorticoid

Receptor

MR Aldosterone,

Deoxy-

corticosterone

Hypertension, Heart

failure

Spironolactone

(Aldactone),

Epleronone

(Inspra)

Introduction __________________________________________________________________________________________________________________________

3

Category

Name

Subtypes/

Isoforms

Natural Ligand

Therapeutic

Relevance

Example of

therapeutic

Ligands

Progesterone Recep-

tor

PR Progestins,

Progesterone

Abortifacient, Men-

strual control

RU486 (Mife-

pristone)

Androgen

Receptor

AR Androgens,

Testosterone

Prostate cancer Flutamide,

Bicalutamide

(Casodex)

Xenobiotic Receptors Constitutive An-

drostane Receptor

PXR Xenobiotics Protection from toxic

metabolites

St. John´s wort,

Rifampicin

Pregnane

Receptor

CAR Xenobiotics Protection from toxic

metabolites

Phenobarbitol

Orphan Receptors Estrogen-related

receptors

ERRα

ERRβ

ERRγ

unknown Muscle fatty metabo-

lism (EERα)

Tamoxifen,

Diethylstilbes-

trol (EERγ)

RAR-related recep-

tors

RORα

RORβ

RORγ

Cholesterol,

Cholesterol

sulfate

Bone maintenance,

circadian rhythm,

cerebellum develop-

ment

-

Human nuclear

receptors 4

HNF4α

HNF4γ

Palmitic acid Role in diabetes -

Reverse erbA Rev-erbAα

Rev-erbAβ

unknown Circadian rhythm -

Testis receptors TR2

TR4

unknown unknown -

Tailless-like TLX

unknown Role in Neuronal

development

-

Photoreceptor-

specific nuclear

receptor

PNR

unknown Role in Photoreceptor

differentiation

-

Chicken ovalbumin

upstream promoter-

transcription factor

COUP-TFI

COUP-TFII

COUP-

TFIII

unknown Role in neuronal

development, vascu-

lar development

-

NGF-induced factor B NUR77

unknown Role in thyomcyte

apoptosis

-

Introduction __________________________________________________________________________________________________________________________

4

Category

Name

Subtypes/

Isoforms

Natural Ligand

Therapeutic

Relevance

Example of

therapeutic

Ligands

Nur-related factor 1 NURR1

unknown Role in dopaminergic

neuron development

-

Neuron-derived

orphan receptor 1

NOR1

unknown unknown -

Steroidogenic factor

1

SF1

Phospolipids Role in sexual devel-

opment

-

Liver receptor ho-

mologous protein 1

LRH1

Phospholipids Role in lipid-

homeostasis, Cell-

cycle control

-

Germ cell nuclear

factor

GCNF

unknown Role in vertebrate

embryogenesis

-

NR-like, DBD-less

recptors

DSS-AHC critica

region on the chro-

mosome gene 1

DAX1

unknown -

Short heterodimer

partner

SHP

unknown -

Table 1. The nuclear receptor superfamily, their natural ligands and therapeutic relevance. (adopted and modified from Moore et al.10)

1.2. Nuclear receptor architecture

Conceptually, the ligand-dependent nuclear receptors bind as trans-acting proteins to

cis-acting DNA sequences (and chromatin) in promoter regions to initiate transcription

of their target genes upon ligand binding. Thus, they integrate endocrine signals to pro-

duce a cellular output. The receptors have a modular structure consisting of six domains

(A to F) including the highly conserved DNA-binding domain (DBD) and the ligand-

binding domain (LBD).12

The modular structure of nuclear receptors is highly conserved in the nuclear receptor

superfamily.

From the N- to the C- terminus the functional domains are as follow (Figure 1):

The A/B domain contains the ligand-independent activation function 1 (AF-1)

that serves as interaction interface for co-activator (CoA) proteins containing a

LXXR motive.

Introduction __________________________________________________________________________________________________________________________

5

The C-domain or DNA binding domain (DBD) is required for the interaction with

specific DNA sequences (half site response elements) within the promoter region

of target genes.

The D region is a hinge that connects the DNA binding domain with the Ligand-

binding domain (LBD) of the receptor.

The E domain or Ligandbinding domain (LBD) contains the ligand-dependent ac-

tivation function 2 (AF-2) with the ligand binding pocket important for the inter-

action with coregulatory proteins (corepressors, CoRs and co-activators, CoAs).

Essentially this domain consists of 12 α-helices that form a hydrophobic

pocket.13-16

The domains C to E form the dimerization interface necessary to form homodimers or

heterodimers with the retinoid X receptor (RXR).

Binding of a ligand to the LBD causes a change of the receptor conformation that chang-

es the affinity to certain coregulatory molecules that may stimulate transcription (co-

activators) or repress transcription (corepressors). This recruitment leads to a remodel-

lig of the chromatin and a modification of the transcriptional machinery.17,18

The structural and functional domains are shown in Figure 1.

Figure 1. Nuclear receptor architecture with their functional domains.

Introduction __________________________________________________________________________________________________________________________

6

1.3. Nuclear receptors as drug targets

Nuclear receptors have a history as drug targets and this has several reasons. They se-

lectively bind drug like molecules and a single-receptor usually regulates a diverse set of

biological functions that have key roles in development, homeostasis and many diseases

such as obesity and diabetes,19,20 cancer21,22 and neurological disorders.23

Nuclear receptor ligands can be full agonists or antagonists and besides that they also

bind partial agonists and antagonists that make them good candidates as drug targets.

Partial (or mixed) ligands bind less efficient and with lower affinity to the receptor and

thus induce a very unique conformational change that may lead to the recruitment of

specific sets of coactivators. These partial ligands are selective modulators and are re-

ferred to as selective NR modulators (SNuRMs). Selective liver X receptor modulators

(SeLRMs), selective peroxisome proliferator-activated receptor modulators (SPARMs)

are examples of this kind of mixed ligands. 24-26 It is believed that such selective com-

pounds act in a more specific way as drug targets overcoming certain side effects.27 The

principle paradigm behind the action of such SNuRMs is that they activate transcription

of only selected target genes in a tissue specific manner to avoid side effects.

There are two critical points that have to be considered in the development of selective

nuclear receptor modulators. First, one needs to identify the coregulators that are asso-

ciated with a certain nuclear receptor and the other critical point is that the selective

activity on certain promoter regions must be given and predictable. The first critical

point in identifying SNuRMs is that the cofactors associated to the certain receptor in a

certain tissue during development must be known. There has been some effort done to

identify coregulators and to determine their tissue distribution and expression

profiles.28 In general there are 50-100 different cofactors associated with different nu-

clear receptors. Some cofactors bind nearly all nuclear receptors (e.g. steroid receptor

co-activator SRC-1) while others are much more specific (e.g. peroxisome proliferator-

activated receptor γ co-activator PGC-1). Some of the coregulators are tissue specific

and/or transiently expressed during development.29,30

Another critical point for the development of SNuRMs is to find ligands that selectively

activate the transcription of specific target genes or act in a tissue specific way to over-

come unwanted secondary actions.31,32

Introduction __________________________________________________________________________________________________________________________

7

1.4. The retinoid X receptor (RXR)

Retinoid X Receptors belong to the steroid/thyroid hormone superfamily designated as

NR2B1-333. There are three different RXR-isoforms: RXRα, RXRβ and RXRγ that are all

encoded by different genes and are differentially expressed in different tissues.

RXRα is mainly expressed in epidermis, intestine, kidney, liver and macrophages. RXRβ

is expressed ubiquitously and RXRγ is mainly expressed in the muscle and parts of the

central nervous system.34,35

1.4.1. Hetero- and homo-dimerization

The Retinoid X Receptor can act as homodimer by the dimerization with itself or may

form heterodimers together with other nuclear receptors such as the retinoic acid re-

ceptors RARα/β, the liver X receptors LXRα/β, the peroxisome proliferators-activated

receptors PPARα/δ/γ, the farnesoid receptor FXR, the thyroid hormone receptor TRα/β

and the vitamin D receptor VDR. There is no obvious preference for a certain RXR iso-

form for most of the heterodimerization partners.5

Furthermore, RXR may form tetramers that are transcriptionally active either in the

presence or absence of a non-activating ligand (a trans-isomer of 9-cis retinoic acid).36

RXR thus builds a unique category in the nuclear receptor superfamily with an im-

portant role in a very diverse set of biological functions and nuclear receptor-mediated

signalling pathways (Table 1).

1.4.2. RXR architecture and transactivation

Retinoid X Receptors have a modular structure to provide a surface for integrating in-

tracellular signals to form a certain output. This concept is reflected in the architecture

of the nuclear receptors that allows essentially receptor-protein, receptor-DNA and re-

ceptor-chromatin interactions induced by ligand binding via allosteric mechanism. The

architecture of nuclear receptors is highly conserved and is sketched in Figure 1.

The main domains are the DNA-binding domain (DBD) and the ligand-binding domain

(LBD). The LBD can be seen as an input-output processor. A certain input (ligand bind-

ing or modifications such as phosphorylation of certain amino acid residues) lead to an

Introduction __________________________________________________________________________________________________________________________

8

allosteric change of the receptor surface that are actual docking sites for the transcrip-

tion machinery, chromatin remodelling complexes that represent the output of the pro-

cessor. Furthermore there are coregulators involved in the regulation of this communi-

cation in the cellular context and the ratio of corepressors (CoRs) and co-activators

(CoAs) determine if a certain ligand acts as agonist or antagonist in a cell-specific or tis-

sue specific manner.32,37

1.4.2.1. The RXR DNA-binding domain

The DNA-binding domain of nuclear receptors is highly conserved in the nuclear recep-

tor superfamily and comprises two zinc finger motives and two α-helices. The human

RXRα DBD (aa130-209) contains the zinc-finger domains at position aa135-155 and

171-190 each complexing a zinc (II) ion through four cysteines. The domain recognizes

selectively DNA regulatory elements that are composed of certain tandem repeats that

are in the case of RXR heterodimers sites containing the sequence AGGTCA arranged in

a direct repeat configuration. Characteristic is that inter-half spacings are 1-5 bp long.

These response elements are known as DR1-DR5, respectively. The direct repeat (DR)

half sites separated by 1-5 base-paires to which the different RXR isoforms bind as ho-

mo- or heterodimers with their respective partner nuclear receptors are summarized in

Table 2. Other members of the nuclear receptor superfamily recognize response ele-

ments containing inverted or everted repeats.38,39

In direct repeats, RXR generally occupies the 5’-element. RXR can also form RXR-RXR

homodimers that bind to DR1.40

Introduction __________________________________________________________________________________________________________________________

9

RXRE Direct Repeats

5´-AGGTCA(n)xAGGTCA-3´

3´-TCCAGT(n)xTCCAGT-5´

Table 2. RXR-DNA-binding domain-DNA interaction. The half sites (5´-AGGTCA-3´) that are recognized

by the different heterodimers or homodimers are listed. They are all direct repeats with 1-5 bp spacings.

N=undefined nucleotide, X= 1-5 base pairs. Sequences that are degenerated do also exist. (adopted from Daw-

son et al. 2012.34)

1.4.2.2. The RXR ligand-binding domain

The RXR ligand binding domain of all nuclear receptors consists of a single protein do-

main. Structural studies have brought more insights how ligand binding within this sin-

gle protein domain leads to the recruitment of certain cofactors. The ligand binding do-

main has a certain tertiary structure that typically consists of α-helices that are arranged

in three layers. This layers form a “sandwich”-like structure to from a hydrophobic lig-

RXR

Response Element

RXR

3´Binding Partner

DR-1 RARα, β, γ

PPARα, β/δ, γ

HNF4α, β

COUP-TFI,II

RXRα, β, γ

DR-2 RARα, β, γ

PPARα, β/δ, γ

RXRα, β, γ

DR-3 VDR

DR-4 TR

LXRα, β

RARα, β, γ

DR-5 RARα, β, γ

TR3/Nur77/NGFI-B

Introduction __________________________________________________________________________________________________________________________

10

and binding pocket for the interaction of the characteristic hydrophobic nuclear recep-

tor ligands.41 The globular structure of the ligand binding domain consists of 11 α-

helices (helix 1-11) whereas helix 12 (H12) works as a mobile arm to the ligand binding

pocket. The position of H12 is critical for interaction of cofactors that contain a so called

NR box or LXXLL motif (L: leucine, X: any amino acid). The overall co-activator surface of

the ligand binding domain is formed by helices 3, 5 and 12 (H3, H5 and H12). 42

The NR box contains a leucine that is a hydrophobic amino acid that allows the interac-

tion with the hydrophobic LBD while a highly conserved lysine (H3) and a highly con-

served glutamic acid (H12) forms a clamp by forming hydrogen bonds of the peptide

backbone of the flanking NR box.43 The allosteric changes of the RXR ligand binding do-

main upon ligand binding can be illustrated by comparing the RXRα apo and RXRα holo

structure. Figure 2 shows a comparison of the two different structures and illustrates

that the biggest conformational changes undergoes helix 12 (H12). The holo RXRα LBD

seals the ligand binding pocket whereas in the apo RXRα LBD helix 12 points away from

the ligand binding pocket. Helix 12 contains amino acid residues important for the re-

cruitment of CoAs and for transcriptional activation.44,45

Introduction __________________________________________________________________________________________________________________________

11

Figure 2. The RXRα apo and RXRα holo ligand-binding domain (LBD) (A) The crystal structure of the

unliganded RXRα apo LBD (PDB entry 1LBD45) (B) The co-crystal structure with a ligand (Docosahexaenoic

acid, DHA) (PDB entry 1MV946) (C) Overly of the RXRα apo and RXRα holo ligand-binding domain. Graphics

adopted from Huber K., University of Innsbruck. 47

1.4.2.3. Permissive and non-permissive nuclear receptors

There are two main classes of RXR heterodimers:

Permissive nuclear receptor heterodimers: LXRs, PPARs, FXR, PXR and CAR

Non-permissive nuclear receptor heterodimers:VDR, TRs9,48

Permissive nuclear receptor heterodimers can be activated by ligands of RXR or ligands

for its permissive heterodimerization partner. They can be furthermore activated by

both ligands or in a cooperative, synergistic way. RXR is an active transcriptional part-

ner in permissive heterodimers. Non-permissive nuclear receptor heterodimers can

A B C

Helix 12 Helix 12

Helix 12

Introduction __________________________________________________________________________________________________________________________

12

only be activated by the heterodimerization partner but not via activation of RXR. In this

case RXR is a silent partner (so called subordination).34,49

The RXR-RAR heterodimer is a special case since it is conditionally permissive. Ligand

binding of RAR activates transcription and allows an RXR ligand to enhance transcrip-

tion thus becoming permissive while being non-permissive in the absence of a RAR ago-

nist. The RXR-RAR heterodimer has been termed “conditionally” permissive since it has

been shown in a few cases that the heterodimer can be activated in the absence of a RAR

ligand.50

1.4.3. RXR isotypes/isoforms and their tissue distribution

The human RXR genes of the three different RXR isoforms are located on chromosome 9,

6 and 1 (bands q34.3, 21.3 and q22-q23, respectively).51

The three different RXR-isoforms: RXRα, RXRβ and RXRγ are encoded by different genes

that contain 10 exons and are differentially expressed in different tissues.52

The RXR promoter is considered to be a housekeeping gene promoter because of its high

G+C content in the 5´ untranslated region. There have been nuclear receptor DR-sites

(DR-0 1, 3 and 4) in the RXR promoter region identified indicating a possible feedback

regulation.53,54

The expression levels also do differ during different developmental stages. The isoform

RXRα has high expression levels in liver, lung, muscle, kidney, epidermis, intestine and

skin and in macrophages. RXRβ is expressed ubiquitously and RXRγ is mainly expressed

in the muscle and parts of the central nervous system.55-57

There is a certain polymorphism in the human population concerning different variants

of RXR isoforms. In a study published in 2009, a RXRβ c.52C

Introduction __________________________________________________________________________________________________________________________

13

1.5. Retinoid X receptor ligands (rexinoids)

Ligands of the Retinoid X receptors are called rexinoid whereas ligands of the retinoid

acid receptor are called retinoids.

Structural ligands for RXR bind to the ligand binding pocket within the ligand-binding

domain of the receptor and alter its conformation. This induces a communication with

the core of the activation function-2 (AF-2) and allows the formation of a binding grove

for co-activators (CoAs).34

In the following RXR ligands will be discussed that are endogenous (section 1.5.1.), that

are present in dietry (section 1.5.2.), synthetic (section 1.5.3. ) or are natural products

(section 1.5.4.).

1.5.1. Endogenous RXR ligands

Endogenous rexinoids and retinoids derive from the vitamin A (retinol) metabolism.

Retinol is absorbed via the blood stream and is taken up in the cell by retinol-binding

protein. Dehydrogenases convert retinol to retinal that is finally converted from retinal

to retinoic acid (RA). RA is bound by cellular retinoic-acid binding proteins (CRABP) that

transport RA to the nucleus in order to act on the respective nuclear receptors.59

9-cis retinoic acid was considered to be an endogenous ligand for RXR but this is contro-

versial, because under normal conditions this vitamin A- metabolite is not detectable in

serum whereas all-trans-retinoic acid and 13-retinoic acid can be detected. Only when

excess all-trans retinoic acid is administered in animal experiments 9-cis retinoic acid is

detectable in serum because of the rapid isomerization from all-trans retinoic acid. 60

However, 9-cis retinoic acid was initially described to be present in the liver and in kid-

ney in higher concentrations compared to all-trans retinoic acid.61

There are other in vivo vitamin A metabolites described that are endogenous ligands for

RXR such as retinal or dehydro-retinoids. Dehydro-retinoids derive from all-trans-

retinol to produce all-trans-13, 14-dihydroretinol that is finally converted to all-trans-

13, 14-dihydroretinoic acid that acts on RXR/RAR heterodimers.62 9-cis-13, 14-

dihydroretinoic acid is the first endogenous ligand for RXR described with a physiologi-

cal relevance in mammals, although the metabolic pathway is unknown.63

Introduction __________________________________________________________________________________________________________________________

14

Figure 3. Structures of selected retinoid X receptor ligands.

9-cis-retinoic acid 9-cis-13,14-dihydroretinoic acid DHA

CD3254

Arachidonic acid

Bexarotene SR11237

Bigelovin Danthron

HX531 Honokiol Magnolol

Rhein

Introduction __________________________________________________________________________________________________________________________

15

1.5.2. RXR ligands from dietary

Dietary- derived ligands for RXR are unsaturated fatty acids such as docosahexaenoic

acid (DHA, 22:6), arachidonic acid (20:4) and oleic acid (18:1) and also other eico-

sanoids were found to activate RXR (Figure 3). These fatty acids are not selective for

RXR but rather activate also other nuclear receptors. 32

Common to these dietary ligands is that they all share a flexible skeleton that fits to the

L-shaped ligand-binding domain of RXR.

The promiscuity of the nuclear receptor RXR to bind different metabolite dietary high-

lights the role of RXR as central sensor of the metabolism and its role in with other nu-

clear receptors such as the PPARs, LXR or FXR to maintain lipid-, bile acid- and glucose

homeostasis in the body.9

1.5.3. Synthetic RXR ligands

A series of RXR ligands have been developed synthetically. One of them is the pan-RXR

full agonist bexarotene (Targretin™, LGD109) that is used to treat subcutaneous T-cell

lymphomas.64,65 The highly conserved ligand-binding domain (LBD) of RXR makes it ra-

ther difficult to develop isoform-selective RXR ligands Furthermore, due to the existence

of permissive heterodimers; synthetic RXR agonists may activate different nuclear re-

ceptor heterodimers and thus possess various pharmacological effects. This may lead to

various side effects as seen in the case of Bexarotene. Observed side effects caused by

bexarotene are high plasma triglyceride levels, hepatomegaly (via activation of LXR) and

suppression of the thyroid hormone axis (via activation of TR).66 In contrast, a hetero-

dimer-selective ligand has been synthesized, LG101506, that activate PPARα and PPARγ

but do not suppress thyroid hormone signaling and is used as insulin sensitizer. Other

selective pan RXR-agonists are CD3254 and SR11237.

Examples of synthetic pan-antagonists are UVI3003 and the dibenzoazepine HX531

(Figure 3).67,68

Introduction __________________________________________________________________________________________________________________________

16

1.5.4. Natural products and derivatives as rexinoids

Natural product derived rexinoids represent a very diverse set of molecules that bind

the nuclear receptor RXR. The fact that such dissimilar structures are able to act as func-

tional rexinoids reflects the conformational adaptability of the RXR ligand binding pock-

et.

The diterpenes hydrophene acid and methoprene acid have been shown to transactivate

RXR in a cell-based transactivation model in Schneider cells. Methoprene acid is a me-

tabolite of microorganisms that metabolize the pesticide methoprene used for mosquito

control. Methoprene itself was designed structurally related to the juvenile hormone of

insects (JHIII) and blocks metamorphosis of these insects. It has been considered that

methoprene is safe although its action on RXR possibly affects retinoic acid- signalling

during development. 69

The natural product phytanic acid has been shown to transactivate RXR and also has

been shown to bind PPARα. Phytanic acid is a metabolite of phytol that originates from

chlorophyll metabolism and can be taken up from the diet. 70,71

Naturally occurring RXR antagonists are β-apocarotenoids (cleavage products from β-

carotene) such as β-apo-14´-carotenal that antagonizes RXR and PPAR activation. Β-apo-

13-caroteneone is a highly potent RXRα antagonist. It has been shown that this natural

compound is able to induce RXR-tetramerization that silences this nuclear receptor.72,73

The sesquiterpene lactone Bigelovin isolated from the flowers of the plant Inulla hu-

pehensis acts as an antagonist of the LXR/RXR heterodimer and on the other hand en-

hances PPARγ/RXR transactivation. This plant is used in traditional Chinese medicine

and it was shown that this natural rexinoid inhibits cell growth of several cancer cell

lines. 74

The napthoquinones danthron and rhein are natural rexinoids isolated form another

plant used in traditional medicine called rhubarb (Dahuang) that derives from the plant

Rheum plamatum. Both rexinoids are specific RXRα antagonists.75

The neolignans honokiol and magnolol are representatives of another naturally occur-

ring compound class that act as rexinoids. Neolignans in general are built up of two

C6C3 units (two propylbenzenes) where the two propylbenzenes are linked at any car-

bon except the β-carbon of the propyl side chain.

Introduction __________________________________________________________________________________________________________________________

17

These natural products derive from the bark of Magnolia obovate or Magnolia officinalis

or other Magnolia species that is used in traditional Japanese medicine (Kampo pre-

scription, Hou Po).76,77

The structures of selected rexinoids are illustrated in Figure 3.

1.5.4.1. Honokiol and magnolol and their pharmacology

Honokiol and magnolol are biphenylic neolignans and thus belong to the polyphenols.

According to IUPAC honokiol is named 2-(4-hydroxy-3-prop-2-enyl-phenyl)-4-prop-2-

enylphenol an magnolol 4-allyl-2-(5-allyl-2-hydroxy-phenyl) phenol.

Honokiol, as well as magnolol, have been initially isolated from the bark of Magnolia of-

ficinalis but can be found in various other Magnolia species.78 Both molecules have two

hydroxyl groups as phenyl side chains and are isomers as shown in Figure 3. Honokiol is

a highly pleiotropic compound and possesses various pharmacological effects and is

thus involved in a lot of different cellular pathways. The compound has main actions in

the cardiovascular system, the central nervous system and the gastrointestinal system

and has been shown to have anti-tumorigenic, anti-inflammatory, and antioxidant ef-

fects.79,80 In cell-based luciferase gene transactivation models it was found that honokiol

activates RXRα with higher potency compared to DHA and phytanic acid but was less

potent in comparison to 9-cis retinoic acid.81 In transactivation models shown by Kotani

et al 2010 honokiol does not activate RARα and β/δ and only weakly PPARγ. In cells it

activates LXR/RXR heterodimers and enhances mRNA levels of the LXR/RXR target

genes ABCA1 and ABCG1. Honokiol furthermore activated cholesterol efflux in perito-

neal macrophages together with the endogenous LXR ligand (22R)-hydroxycholesterol

in a synergistic way. Taken together, in the cellular context honokiol acts via the modu-

lation of the LXR/RXR heterodimer.82 Furthermore, honokiol has been described as very

weak partial PPARγ agonist with non-adipogenic properties. In contrast, magnolol is a

more potent PPARγ agonist in comparison to honokiol.83 The neuro-modulatory effects

of Magnolia bark extract are partly due to the action of honokiol (and magnolol) on the

GABAA receptor. It is believed that it acts similar to benzodiazepines but with higher

subunit selectivity.84, 85 In another study it has been shown that honokiol affects the syn-

thesis of GABA itself. 86

Introduction __________________________________________________________________________________________________________________________

18

Additionally, honokiol and magnolol are acting on cannabinoid receptors (cannabinoid

receptors 1 and 2, CBR1 and CBR2). Magnolol is a partial CBR2 agonist, whereas

honokiol is a CBR2 antagonist and a full CBR1 agonist. 87 Due to their properties to act

on the GABAA and CB receptors honokiol and magnolol have been used in previous stud-

ies as lead structure for the design of honokiol derivatives with more selective and more

effective properties for these receptors. 84,88-90

1.5.4.2. Partial agonists of RXR

Partial agonists activate the receptor less efficient in comparison to full agonists and are

sometimes considered to display both agonistic and antagonistic effects at the same time

resulting in a decrease of overall efficacy.91

A study from De Lera and Bourget 2007 addressed the question of how partial agonism

may work in the RXR LBD. They compared the co-crystal structures with the full agonist

CD3254 and with CD3254-derivatives (partial agonists) to reveal the amino acid side

chains involved in the agonist-partial agonist-to antagonist transition.

They quantified the impact of ligand binding on the motion on helix 12 (H12) and found

that the difference was a certain amino acid side chain in helix 11 (L436) that was af-

fected. They showed that the interaction with this residue caused by partial agonists

leads to a destabilisation of the holo conformation of H12 in solution that causes a de-

crease in the interaction with coactivators. Thus, full agonists in comparison to partial

differ in their impact to stabilize holo-H12. 92 Therefore, what matters is how a ligand

“senses” the intracellular co-regulator levels and thus may act as tissue selective modu-

lator and may, depending on the context act as agonist or antagonist.93

It is hypothesized that partial RXR agonists (and antagonists) have a therapeutic poten-

tial since they display less side effects in comparison to full agonists.94 This therapeutic

potential is to date largely unexplored. In a study CBT-PMN, a partial selective RXR ago-

nist, was investigated and compared to full agonists in mice in the context of type 2 dia-

betes and has been shown to produce less side effects in a mouse model of Type 2 diabe-

tes in comparison to a full agonist (no increase of serum triglyceride levels, body weight

and cholesterol levels in the blood).95

Introduction __________________________________________________________________________________________________________________________

19

1.5.4.3. Rexinoids with heterodimer selectivity

Rexinoids that are selective for certain heterodimers are called selective RXR modula-

tors (specific NR modulators, SNuRMs) and built an important class of ligands.

The selectivity and functional consequence of a ligand is determined by the conforma-

tional change induced in the LBD of the receptor. This conformational change lead to the

recruitment or repulsion of different coregulators that interact with the LBD to regulate

gene expression in a cell-type specific manner.

This class of compounds may act as agonists or partial agonists of selective permissive

heterodimers or as synergists where they increase potency for the partner ligand or as

selective antagonists where they can inhibit synergistic RXR agonist activity in permis-

sive heterodimers.96

Table 3 summarizes some rexinoids that have been shown to display heterodimer selec-

tivity.97

Selective RXR modulator

Heterodimer selectivity

References

LG100268

LXR/RXR agonist

PPARγ/RXR agonist

98,99

LG101305

AR/RXR antagonist 100

LG101506 PPARγ/RXR agonist

RXR/RXR partial agonist

101,102

L100754 PPARγ/RXR agonist

RXR/RXR antagonist

103

PA024 LXR/RXR agonist

PPAR/RXR agonist

104

HX630 PPARγ/RXR agonist

105

Table 3. Heterodimer selective RXR modulators.

Introduction __________________________________________________________________________________________________________________________

20

1.6. Nuclear receptor coregulators Coregulator proteins function as adaptors to bridge the communication of the nuclear

receptor to the transcriptional machinery and mediate the regulation of transcription.

The first nuclear receptor coregulators were identified in the 90ies in protein-protein

interaction studies. They were classified as co-activators, such as SCR1and corepressor

proteins, such as SMRT and NCoR. 106-108 A high number of coregulator proteins have

been identified in the meantime and more than 350 of them are now listed in the “Nu-

clear Receptor Signalling Atlas” (www.NURSA.org).109

The presence or the absence of a certain ligand determines the coregulators recruited to

the nuclear receptor and thus the repression or activation of transcription in a certain

tissue. Most coactivators do this via a certain motif with the sequence LxxLL (NR box,

see section 1.2.).110 An analogous motif has been identified in corepressor proteins and

is termed CoRNR box (LxxH/IxxI/L).111,112 Which coregulator is recruited to the nuclear

receptor is determined by the position of helix 12 (H12) in the ligand-binding domain of

the respective nuclear receptor.29 The initial proposed mechanism was that the agonist

bound holo-receptor interacts with coactivators that possess a specific enzyme activity

e.g. histone acetyl transferase activity (HAT activity) or possess a chromatin remodelling

function to activate transcription. The non-ligand bound receptor (apo-receptor) inter-

acts with corepressors, e.g. histone deactelyases (HDAC activity) to prevent transcrip-

tion of target genes. However, some corepressors were described that are ligand-

dependent. Furthermore, in some cases the apo-receptor can be modified by post-

translational modifications to activate target gene transcription.113-115

The nuclear receptor coactivator 1 (NCOA1/SRC-1) is a coactivator possessing HAT ac-

tivity and belongs to the p160/steroid receptor coactivator (SRC) family of coregula-

tors.116 Nuclear receptor coactivator 2 (NCOA2/SRC-2/TIF-2/GRIP-1/p160) and nuclear

receptor coactivator 3 (NCOA3/SRC-3/RAC3/ACTR/pCIP/AIB-1) also belong to the SCR-

family of coregulators that possess HAT activity.117

There are functional homologues of these proteins such as CREBP/CBP (CREB binding

protein) and EP300/p300 that have HAT activity and have been shown to interact with

SRC-1, RNA polymerase II and other transcription factors. 118

The peroxisome proliferator-activated receptor gamma coactivator-1 (PGC-1) is a coac-

tivator that modulates RNA processing and is involved in energy metabolism. 119

http://www.nursa.org/

Introduction __________________________________________________________________________________________________________________________

21

Other coregulators are associated with the “mediator” complex such as MED1

(TRAP220/TRIP2/CRSP200) and function as coordinators of transcription factors and

the basal transcriptional machinery. MED1 plays a crucial role in ligand-dependent

chromatin remodelling and may be able to form loops within the DNA to recruit the ma-

chinery to enhancer sequences.120

TBP is a TATA-binding protein that serves as basal activator of transcription.

Together with TAFs (TBP-associated factors), they form the TFIID complex that binds to

the core promoter and helps positioning the polymerase and acts as a scaffold for the

transcriptional complex.121

Main coregulators that interact with RXR are listed in Table 4.

Coregulator Name

Activity

Specific

AF-2 dependent

References

NCOA1

Co-activator

No

Yes

106,122

NCOA2 Co-activator No Yes 106,123,124

NCOA3 Co-activator No - 125,126

PGC1A Co-activator No Yes 119

MED1 Co-activator No Yes 125,127

TBP Co-activator No Yes 125,128

TAF4 Co-activator No Yes 125,128,129

TAF11 Co-activator No Yes 129

CREBP Co-activator No Yes 130,131

EP300 Co-activator No Yes 130

Table 4. Main coregulators described for the retinoid X receptor.

http://www.guidetopharmacology.org/GRAC/CoregulatorDisplayForward?coregId=46http://www.guidetopharmacology.org/GRAC/CoregulatorDisplayForward?coregId=47http://www.guidetopharmacology.org/GRAC/CoregulatorDisplayForward?coregId=48http://www.guidetopharmacology.org/GRAC/CoregulatorDisplayForward?coregId=62http://www.guidetopharmacology.org/GRAC/CoregulatorDisplayForward?coregId=45http://www.guidetopharmacology.org/GRAC/CoregulatorDisplayForward?coregId=77http://www.guidetopharmacology.org/GRAC/CoregulatorDisplayForward?coregId=76http://www.guidetopharmacology.org/GRAC/CoregulatorDisplayForward?coregId=31http://www.guidetopharmacology.org/GRAC/CoregulatorDisplayForward?coregId=35

Introduction __________________________________________________________________________________________________________________________

22

2. Targeting retinoid X receptors in human disease

In the last years effort has been made to develop compounds targeting these nuclear

signal integrators for therapeutic applications such as for the treatment of cancer, meta-

bolic diseases or neurological disorders.65,132

A synthetic RXR ligand, the full RXR agonist Bexarotene (Targretin™, LGD109) is used to

treat subcutaneous T-cell lymphomas and breast cancer but patients experienced severe

side effects when the synthetic drug went through the clinical trials.133,134

Other rexinoids are being tested in preclinical settings for the treatment of atherosclero-

sis or insulin resistance.135

Overall, full retinoid X receptor agonists have been shown to induce hepatomegaly, to

increase plasma triglyceride levels and to suppress the thyroid hormone axis. 136

2.1. The role of nuclear receptors in macrophage biology

Macrophages play a major role in the innate immune system and have main roles in the

first defence mechanisms against pathogens, the clearance of cellular debris, in tissue

remodelling and also in the integration of lipid metabolism in different tissues.137

Furthermore macrophages play a critical role in diseases such as atherosclerosis, insulin

resistance and neurologic disorders.138-140

The isoform RXRα is highly expressed in human blood monocytes and dendritic cells

and RXRβ is expressed at lower levels in this cell types.

Furthermore blood monocytes express low levels of PPARα and moderate levels of

PPARβ/δ, RARα and γ, the LXRs, VDR, Nur77 and Nurr1. Human dendritic cells express

PPARβ/δ and PPARγ, RARα, the LXRs and VDR.141

Retinoid receptors play a major role in monocyte-macrophage differentiation and RXR

target genes are critical for the maintenance of the homeostasis between self-renewal

and differentiation. Furthermore, it has been shown that RXR controls differentiation

and apoptosis in hematopoietic stem cells pointing out how important RXR function is

for myeloid cell fates during the different stages of maturation.141

Heterodimerization with RAR and PPARγ controls self-renewal in the most primitive

hematopoietic stem cells. RAR/RXR inactivation maintains self-renewal and PPARγ/RXR

activation promotes differentiation of these cells. Selective RXR modulators can activate

Introduction __________________________________________________________________________________________________________________________

23

different pathways that have been considered to treat certain pathological conditions

such as myelodysplastic syndromes or atherosclerosis.142

In the context of immune function, the most important nuclear receptors are the PPARs

and LXRs. However, RAR, VDR, PXR and FXR as well as Nur1 and Nur77 are involved

mediating macrophage activation.143-147

Animal studies have shown the beneficial and clinically important effects of RXR ago-

nists in models for chronic inflammatory diseases, such as atherosclerosis, insulin-

resistant diabetes and neurodegeneration. 138-140

2.2. Nuclear receptors in lipid and cholesterol metabolism

Macrophages regulate lipid-metabolism and are crucial for lipid homeostasis by up tak-

ing, storing and oxidizing lipids and executing cholesterol efflux. These mechanisms are

crucial to prevent diseases and keep the homeostatic balance of lipid metabolism up-

right.148

Lipid-ligand activated transcription factors together with RXR regulate storage and re-

lease and elimination of lipids in human macrophages. The main receptors that are in-

volved are the permissive heterodimer receptors PPARs and LXRs.

More recently, PXR, FXR, Nurr1 and Nur77 have been described to also play a role in

energy metabolism but will not be described further here.149-151

2.2.1. The peroxisome proliferator-activated receptors (PPARs)

The three receptor subtypes of PPAR, PPARα, PPARβ and PPARγ (NRC1C1-NR1C3) are

all encoded by different genes and also display distinct biological functions. Together

with RXR as their heterodimer partner, they recognize DR-1 or DR-2 type recognition

sequences within the promoter regions of their target genes (Table 2).

PPARα is expressed in tissues that are of high relevance for lipid and fatty acid catabo-

lism, such as the liver, kidney, skeletal muscle and heart, brown fat and the intestine. 152

The action of this isoform reduces triglyceride levels and low-density lipoprotein levels

(LDL) in the blood and elevates levels of high-density lipoprotein particles (HDL).153,154

Endogenous ligands of PPARα are saturated and unsaturated fatty acids such as arachi-

donic acid, palmitic acid, oleic acid or linoleic acid and important synthetic agonists are

Introduction __________________________________________________________________________________________________________________________

24

fibrates for the treatment of insulin sensitivity, to improve blood glucose levels or hy-

pertriglyceridemia. 26,155,156

PPARβ/δ is also more broadly expressed and is found mainly in the brain, adipose tissue

and the skin. The activity of this nuclear receptor subtype ameliorates glucose and lipid

metabolism primarily in adipose tissue, heart and skeletal muscles.157

PPARγ is mainly expressed in adipocytes and the liver and at lower levels in skeletal

muscles and the liver. Furthermore PPARα and PPARγ are expressed in mono-

cytes/macrophages and vascular wall cells.158,159

Similar to PPARα, PPARγ plays a crucial role in lipid homeostasis as well as in glucose

homeostasis. Moreover, PPARγ has major roles during the inflammatory response.160,161

2.2.2. The liver X receptors (LXRs)

The Liver X Receptors are another subclass that forms heterodimers with RXR.

LXRs exist in two different isoforms (LXRα and LXRβ termed NR1H3 and NR1H2, re-

spectively) and the LXR/RXR heterodimer recognizes specific DNA sequences (DR-4) in

the promoter and regulatory regions of target genes (Table 2).

The un-liganded LXR/RXR heterodimer binds constitutively to this recognition sequenc-

es in a repressed state due to interactions with corepressors that block transcription

and recruit histone deactelyases.162

Endogenous ligands of the LXR/RXR heterodimer are oxysterols and intermediate me-

tabolites of cholesterol biosynthesis. 163,164

Ligand activation of LXR leads to the recruitment of specific co-activators (e.g. SRC-1)

and the interaction of histone acetyl transferases to initiate transcription of target

genes.164 Besides genes important for maintaining cholesterol and lipid homeostasis

LXR/RXR activation inhibits the transcription of certain pro-inflammatory genes.165,166

2.2.3. The retinoid X receptors (RXRs)

The retinoid X receptor controls cholesterol uptake, its efflux and cholesterol storage

mainly due to the activation of permissive heterodimers (see section 1.4.2.3.).

Uptake of lipoproteins is regulated via scavenger receptors. Activation of RXR in macro-

phages with 9-cis retinoic acid or other rexinoids upregulates the expression of the

Introduction __________________________________________________________________________________________________________________________

25

scavenger receptor CD36,167 while other scavenger receptors are downregulated (SRA-

II/II) that overall leads to an decreased lipid storage within cells.167,168 It has been

shown that CD36 regulation in macrophages is mediated by several RXR-heterodimers,

such as PPAR/RXR, RAR/RXR and FXR/RXR.169,170

The regulation of SRA has been shown to be regulated by the permissive action of

Nurr1/RXR and Nur77/RXR heterodimers.171

The efflux of cholesterol out of cells is a highly important mechanism to get rid of excess

cholesterol. The efflux of cholesterol is mediated by the action of different ABC trans-

porters.

Ligand activation of RXR by 9-cis retinoic acid, bexarotene or other rexinoids like

LG100268 and HX630 or the natural occurring rexinoid honokiol have been shown to

promote ABCA1 and ABCG1 expression from different cell lines, including human mac-

rophage cell lines or primary mouse macrophages.167,168,172,173

Other RXR targets important for cholesterol efflux is the cholesterol transport protein

ADP-ribosylation factor-like 7 (ARL4C) and a sterol eliminating enzyme CYP27A1.

The expression of these two proteins is mediated by the permissive action of

PPARα/RXR, PPARγ/RXR and/or the LXRs/RXR.167

Furthermore, RXR activation is involved in the processing as well as the storage of lipids.

In human macrophages, RXR activation leads to the induction of the expression of

apolipoprotein E (ApoE) that promotes the efflux of lipids to apolipoproteins.

Another key regulator in cholesterol and lipid homeostasis is the transcription factor

SREBP1 that induces a lipogenic gene program (e.g. induction of fatty acid synthase, ace-

tyl-CoA-carboxylase) and leads to the transcription of genes involved in cholesterol bio-

synthesis and uptake. SREBP1 is a target of the LXR/RXR heterodimers.174

Introduction __________________________________________________________________________________________________________________________

26

2.3 Atherosclerosis, lipid- and cholesterol metabolism

Atherosclerosis is a chronic inflammatory disease hallmarked by thickening and harden-

ing of artery walls. The disease causes coronary and cerebrovascular diseases that are

the two major morbidities worldwide.175

There are a lot of different environmental and genetic risk factors such as hypertension,

obesity, insulin resistance or type 2 diabetes.176 However, the disease is characterized by

a local immune response that is caused by a deposition of cholesterol, lipid and cellular

debris followed by a deposition of white blood cells into the sub endothelium. These

depositions may become persistent and are then called atherosclerotic plaques.138

Nuclear receptors are heavily involved in lipid and cholesterol homeostasis as well in

glucose homeostasis and the role of these receptors in the control of macrophage gene

regulation and gene expression has been therefore studied extensively during the past

years.167,177 Indeed, the main constituent of such depositions are cholesterol and lipids

and consequently atherogenesis occurs by the infiltration of the sub endothelial space

by monocytes that subsequently start to differentiate into macrophages. Macrophages

express lipoprotein receptors or scavenger receptors that mediate the uptake of lipids

and cholesterol from these deposits. If the lipid accumulations exceed a critical point, the

accumulation of these lipids in macrophages leads to the formation of foam cells that

drives the lipid deposition further. Thus, macrophages are important cholesterol-

accumulating cells in atherosclerosis and efflux of cholesterol out of these lipid-laden

cells under pathologic conditions might protect against atherosclerosis.178

Reverse cholesterol transport (RCT) is the major physiological process that promotes

cholesterol efflux from peripheral tissues by high-density lipoprotein (HDL) to deliver

the excess cholesterol to the liver. From the liver cholesterol is partly eliminated

through the bile and the feces. This process maintains the cholesterol homeostasis by

maintaining the balance of cholesterol intake and de novo cholesterol synthesis.

In macrophages, multiple genes regulate RCT.

Modified lipoproteins are taken up by scavenger receptors in macrophages and are de-

livered to endosomes/lysosomes where the initial step of RCT occurs: the hydrolysis of

cholesterol esters (CE) to free cholesterol (FC) and fatty acids. This initial step is medi-

ated by lysosome acid lipase (LAL). Processed cholesterol is then integrated into the cel-

lular membrane. Niemann Pick type C 1 and 2 proteins (NPC1 and NPC1) control these

Introduction __________________________________________________________________________________________________________________________

27

intracellular mechanisms.179 Excess cholesterol on the other hand is transported to the

endoplasmic reticulum and is there re-esterified with fatty acids by acyl-CoA cholesterol

acyltransferase 1 (ACAT1) and is stored as lipid droplets.180

Three main transporters play a critical role in macrophage cholesterol efflux. ATP-

binding cassette transporter ABCA1, ABCG1, ABCG4 and the scavenger receptor type

I(SR-BI). The main cholesterol transporters are expressed under the regulation of the

action of RXR and LXR, mainly the permissive heterodimer LXR/RXR. 181 The transport-

ers interact with the cholesterol acceptors HDL and Apo-AI and effluxed cholesterol is

then carried out by HDL to the liver to be taken up by the liver SR-BI (termed direct

pathway). Alternatively, free cholesterol is transferred by CEPT (cholesteryl ester trans-

fer protein) to apoB-containing lipoproteins that are later cleared by the liver via recep-

tors such as low-density lipoprotein receptor (LDLR) (termed indirect pathway). 178

Under atherosclerotic conditions the lipid accumulation in macrophages take overhand,

that lead to the formation of foam cells that remarkably influences the progress of the

disease progression.182

For this reason, one strategy to treat atherosclerosis and prevent atherosclerotic genesis

would be to enhance efflux of excess cholesterol out of peripheral cells(macrophages) to

prevent cellular cholesterol retention by pharmacotherapy.183,184

2.3.1. Ligands for RXR in lipid-handling related diseases

It has been shown that rexinoids significantly reduces the progression and development

of atherosclerosis in mouse models of dyslipidaemia and they do so by enhancing the

capacity of cholesterol efflux within macrophages. 170,185

Ligands for PPARs and LXR are reported to possess similar effects in disease models that

indicate that the effect of rexinoids might be due to these receptor heterodimers in

vivo.132,145

Rexinoids have been reported to have potential in the treatment of other lipid-handling

related diseases as well, such as certain neurological conditions.186,187

Bexarotene was reported in 2012 by Cramer et al to clear β-amyloid plaques and im-

prove cognitive deficits in a mouse model of Alzheimer’s disease (AD).188

One hypothesis is that the therapeutic effect in AD models is due to the upregulation of

ApoE and the increase in ABCA1 expression mediated by the activation of LXR/RXR and

PPAR/RXR permissive heterodimers.189

Introduction __________________________________________________________________________________________________________________________

28

Finally, ligands for retinoid X receptors have a therapeutic potential for treating lipid-

related and metabolic diseases but their development is still challenging because of off-

target effects of permissive RXR heterodimers, cytotoxicity and tissue availability of

RXR. Observed side effects of activation of RXR is rising of plasma triglyceride levels and

the suppression of the thyroid axis via the activation of permissive RXR

heterodimers.190,191 Therefore, development of novel RXR-selective ligands for therapy

aims to develop heterodimer selectivity with improved action.192

Introduction __________________________________________________________________________________________________________________________

29

3. Honokiol/magnolol derivatives used in this study

Semisynthetic honokiol derivatives synthesized at the Institute of Applied Synthetic

Chemistry (Univ.Prof. Dipl.-Ing. Dr.techn. Marko Mihovilovic, Vienna University of Tech-

nology, Vienna)193 and the Institute of Pharmaceutical Sciences (Priv. Doz. Dr. Wolfgang

M. Schuehly, Karl-Franzens-University Graz, Graz)84,194,195 were initially screened in a

cell-based luciferase-reporter gene assay in HEK293 and HepG2 cells for nuclear recep-

tor activation in vitro (see section 2.3.2.-2.3.5).

In total 55 neolignans were tested for transactivation of RXRα, the PPAR isoforms α, β/δ

and γ, the LXR isoforms α and β and FXR. The screening was performed by Dr. Angela

Ladurner, Dr. Clemens Malainer and former diploma students from the Department of

Pharmacognosy, University of Vienna (Reem Selim, Amina Cocic and Andrea Holzer) and

by myself.196-198

The structures of tested semisynthetic neolignans and the results of the screening are

displayed and summarized in the Appendix (Appendix Figure 1 and Table 1).

Five derivatives selectively activated human full length RXRα in this cell-based luciferase

testing model.

Two semisynthetic honokiol derivatives from the Institute of Applied Synthetic Chemis-

try (Group of Univ.Prof. Dipl.-Ing. Dr.techn.Marko Mihovilovic, University of Technology,

Vienna) were finally selected for further functional characterization.

Compounds 2284 (LRK017) and 2817 (LRK363) were synthesized by Dr. Lukas Rycek

and Dr. Dominik Dreier at the Vienna University of Technology.199

Introduction

30

31

Aim of the work

Considering the important biological role of the retinoid X receptor the aim of this work

was to identify new partial agonists for the retinoid X receptor that act as selective mod-

ulators structurally based on the natural compounds and naturally occurring rexinoids

honokiol and magnolol.

Potent compounds that responded in an initial screen in a cellular luciferase-based test-

ing model in HEK293 and HepG2 cells in a dose-dependent manner were selected and

further investigated.

A coregulator interaction profile of the RXRα ligand-binding domain (LBD) of 154 pep-

tides representing 64 possible interacting coregulators was performed and compared to

co-regulator interaction profiles of full RXR agonists

In vitro binding studies and functional studies of potent RXRα agonists further lead to a

pharmacological and bioactivity profile of active honokiol derivatives.

.

Materials and Methods

32

Material and Methods

1. Materials

1.1. Products and supplier information Name Supplier

Cell Culture Material Benzylpenicillin

Lonza Group Ltd. (Basel, Switzerland)

Dulbecco´s Modified Eagle Medium (DMEM)

Lonza Group Ltd. (Basel, Switzerland)

Foetal bovine serum Lonza Group Ltd. (Basel, Switzerland) L-glutamine Lonza Group Ltd. (Basel, Switzerland) Penicillin (potassium salt)-Streptomycin sulphate

Lonza Group Ltd. (Basel, Switzerland)

Trypsin Invitrogen (CA, USA) Biologicals and Chemicals Amphicillin (sodium salt) Adenosin-5´triphosphate disodium salt Agar ApoA1 recombinant Bodipy®493/503 Bovine serum albumine Coenzyme A (trilithium salt) CompleteTM Dithiothreitol (DTT) D-Luciferin (sodium salt) Dexmethasone DMSO Insulin IBMX Kanamycin LB broth Lipofectamine®2000 Oil red O PMA p-coumaric acid PMSF Reporter lysis 5X buffer Resazurin (sodium salt) Roti®Quant TEMED Tritium labeled Cholesterol

Sigma (MO, USA) Sigma (MO, USA) Sigma (MO, USA) Sigma (MO, USA) Molecular Probes (O, US) New England Biolabs (MA, USA) Sigma (MO, USA) Roche Diagnostics, (Penzberg,Germany) Fluka (MO, USA) Synchem (Flesberg, Germany) Sigma (MO, USA) Fluka (MO, USA) Sigma (MO, USA) Sigma (MO, USA) Sigma (MO, USA) Sigma (MO, USA) Invitrogen (CA, USA) Fluka (MO, USA) Sigma (MO, USA) Sigma (MO, USA) Sigma (MO, USA) Promega (Wi, USA) Sigma (MO, USA) Carl Roth (Karlsruhe, Germany) Fluka (MO, USA) PerkinElmer (MA, USA)

Materials and Methods

33

Triton® X-100 Unesterified Cholesterol (UC) [1,2-3H(N)]-Cholesterol 49.0 Ci/mmol

Sigma (MO, USA) Sigma (MO, USA) Perkin Elmer (Boston, MA, USA)

Compounds 9-cis retinoic acid Bexarotene GW3965 HX351 Honokiol Magnolol Pioglitazone Rosiglitazone T0901317

Cayman (MI,USA) Sigma (MO, USA) Sigma (MO,USA) Tocris Bioscience (Washington DC, USA) University of Vienna, J.Rollinger (Austria) University of Vienna, J. Rollinger (Austria) Molekula (Shaftesbury, UK) Cayman (MI,USA) Sigma (MO, USA)

Table 5. Cell culture material, biologicals and chemicals.

1.2. Cell culture media and supplements Name Components Amount HEK293 and HepG2 growth medium (DMEM complete)

DMEM high glucose (phenol red free) Penicillin (potassium salt) Streptomycin sulphate L-glutamine Foetal bovine serum (heat inactivated)

500mL 100 U/mL 100 µg/mL 2 mM 10 %

DMEMstripped medium FBSstripped

DMEM high glucose (phenol red free) Penicillin (potassium salt) Streptomycin sulphate L-glutamine Foetal bovine serum, stripped (FBSstripped) Foetal bovine serum (heat inactivat-ed) Charcoal powder T70-Dextran powder Sucrose powder HEPES buffer stock pH 7.6 MgCl2 stock

500mL 100 U/mL 100 µg/mL 2 mM 10 % 0.0025% w/v 0.25 M 1M 1M

Materials and Methods

34

Name Components Amount DMEMstarvation medium

DMEM high glucose (phenol red free) L-glutamine

500mL 2 mM

THP-1 medium (RPMI complete)

RPMI high glucose (phenol red) Penicillin (potassium salt) Streptomycin sulphate L-glutamine Foetal bovine serum (heat inacti-vated)

500 mL 100 U/mL 100 µg/mL 2 mM 10 %

3T3-L1 medium (Adipocte growth me-dium)

DMEM high glucose (phenol red free) Penicillin (potassium salt) Streptomycin sulphate L-glutamine Calf serum (heat inactivated)

500 mL 100 U/mL 100 µg/mL 2 mM 10 %

Medium for Luciferase Assays

DMEM (high glucose) L-glutamine

500 mL 2 mM

Adipocyte differentiati-on medium

DMEM high glucose (phenol red free) Penicillin (potassium salt) Streptomycin sulphate L-glutamine Foetal bovine serum (heat inactivated) Insulin Dexamethasone IBMX

500mL 100 U/mL 100 µg/mL 2 mM 10 % 1 µg/mL 250 nM 250 µM

Adipocyte medium DMEM high glucose (phenol red

free) Penicillin (potassium salt) Streptomycin sulphate L-glutamine Foetal bovine serum (heat inactivated) Insulin

500mL 100 U/mL 100 µg/mL 2 mM 10 % 1 µg/mL

Table 6. Cell culture media and supplements.

Materials and Methods

35

1.3. Commercially available kits

Name Supplier LanthaScreen™ TR- FRET RXRalpha Coac-tivator Assay Kit, goat

Invitrogen (CA, USA) (Catalogue number PV4797)

pegGOLD Plasmid Midiprep Kit PeqLab (Erlangen, Germany)

PureYield™ Plasmid Midiprep System Promega (Wi, USA)

RNA Isolation Kit PeqLab (Erlangen, Germany) Superscript™ First-Strand Synthesis System

Invitrogen (CA, USA)

qPCR Green Master Mix Roche Diagnostics, (Penz-

berg,Germany)

Table 7. Commericailly available kits.

1.4. Reagents and buffers Name and Components Amount Solutions for transient transfection: (Calcium phosphate method)

Calcium phosphate- NaHPO4 (stock solution) 5.2 g in 500mL H2O CaCl2 2M 2 x HBS 8.0 g NaCl, 6.5 g HEPES, 10 mL Na-

HPO4 stock solution pH adjusted to 7.4 with diluted NaOH or HCL and filles up to 500 mL

Luciferase Measurement: Luciferase lysis buffer

1.2 mL 5 x Reporter lysis buffer (Promega) 4.8 mL ddH20 6 µl 270 mM Coenzyme A [270 µM] 6 µl 1M DTT [10.1.mM]

Materials and Methods

36

Luciferin Luciferin Tricine

32 mg/mL 21 mM, pH 7.8

ATP Tricine pH 7.8 MgCl2 21.5 mM Adenosin-5´triphosphate disodium salt

20 mM, pH 7.8 21.5 mM 3.8 mM

Cell Viability Assay: 10x Resazurin stock

1 mg/ml in sterile PBS Prior to use, the stock was diluted in sterile PBS 1:10 and further diluted 1:10 with the growth medium in the well plates (DMEM or RPMI + 1% Glutmine)

Western blot solutions and buffers: Protein extraction stock solution RIPA lysis buffer 50mM

Protein quantifcation

50 mM Tris/HCl pH 7.4 500 mM NaCl 5.0 mM NP40 12.06 mM Na-Deoxycholate 3.47 mM SDS 7.7 mM NaN3 dissolved in H2O freshly added prior to use: 4 % Complete (protease inhibitor cocktail, Roche Diagnostics, Penzberg Germany) 1 mM PMSF 1 mM NaF 1 mM NaVO3 1 : 4.75 Roti® -Quant (Roth, Karslruhe, Germany) in Aqua dest

Immunoblotting: Western blot sample buffer 3x

187.5 mM Tris/HCl pH 6.8 0.2 M SDS 30% Glycerol 0.2 mM Bromphenol blue 15% β-Mercaptoethanol in 1000 mL ddH2O

Materials and Methods

37

Resolving gel Stacking gel SDS Running buffer 10x Blotting Buffer 5x TBS-T pH 8.0 10x ECL-solution Solutions for Confocal microcopy: Formaldehyde solution 3% BSA/PBS solution (0.2%) Triton solution (0.2%)

7.5-10 % PAA (Rotiphorese® Gel30) 375 mM Tris/HCl pH 8.8 0.1% SDS 0.1% TEMED 0.05% APS 7.5 mL ddH2O 5 % PAA (Rotiphorese® Gel30) 125 mM Tris/HCl pH 6.8 0.1% SDS 0.1% TEMED 0.05% APS 3.75 mL ddH2O 248 mM Tris-base 1.9 M Glycine 35 mM SDS in 1000 mL ddH2O 125 mM Tris-base 971 mM Glycine in 1000 mL H2O 248 mM Tris-base 1.9 M NaCl 1 % Tween-20 in 1000 mL H2O 100mM Tris-base pH 8.5 1.24 mM Luminl 0.2 mM p-Coumaric acid 0.018% H2O2 In 10 mL H2O Formaldehyde stock solution (37%) 33ml, PBS solution 374 ml 3g BSA in 1500 mL PBS solution 3 g 1500 ml (mixed overnight and sterile-filtered) 800 µl Triton in 400 ml BSA/PBS so-lution (0.2%) (mixed overnight and sterile-filtered)

Table 8. Reagents and buffers.

Materials and Methods

38

1.5. Plasmid DNA Name of plasmid Source Antibiotic

resistance

Used restriction

enzymes

Expression plasmids

EGFP

(pEGFP-N1)

Clontech (Mountain View,

CA, USA)

Kanamycin Nhell, EcoRI

Human RXRα,

full length

(pcDNA3.1+)

Missouri S&T cDNA Res

(Rolla, Missouri, USA)

Ampicillin EcoRI, Xhol

Human LXRα,

full length

(pCMV)

Missouri S&T cDNA Res

(Rolla, Missouri, USA)

Ampicillin ApaI, KpnI

Human LXRβ,

full length

(pcDNA3.1+)

Missouri S&T cDNA Res

(Rolla, Missouri, USA)

Ampicillin ApaI, KpnI

Human PPARγ

Full length

(pSG5-PL-hPPAR γ1)

Univ. Prof. Walter Wahli

and Univ. Prof. Beatrice

Desvergne

Ampicillin EcoRI, HindIII

GAL4 hRXRαLBD

Chimeric Expression of

GAL4-LBD (pCMX)

Univ. Prof. Ronald M. Ev-

ans (Howard Hughes Medi-

cal Institute, La Jolla, CA)

Ampicillin Functionally vary-

fied

GAL4 hPPARγ LBD

Chimeric Expression of

GAL4-LBD (pCMX)

Univ. Prof. Ronald M. Ev-

ans (Howard Hughes Medi-

cal Institute, La Jolla, CA)

Ampicillin

unctionally varyfied

pET15b-His6 human-

RXRα

(pET15b)

Dr. Marcel Scheepstra

(Technische Universiteit,

Eindhoven, The Nether-

lands)

Ampicillin Protein

sequenced

Reporterplasmids

RXR_RE

(RXR(2) Luciferase Re-

porter Vector)

(pTL Luc)

Missouri S&T cDNA Res

(Rolla, Missouri, USA)

Ampicillin HindIII, BamHI

Materials and Methods

39

Name of plasmid Source Antibiotic

resistance

Used restriction

enzymes

ABCA1promoter

-703/-38

Luciferase Reporter Vec-

tor

(pGL4.14)

Dr. Irena Ignatova (University of Virginia

Health System, Char-

lottesville, Virginia, USA)

Ampicillin

KpnI, SacI

SREBP1c promoter

- 556/+42

Luciferase Reporter Vec-

tor

(pGL3)

Dr. Gyun Sik Oh

(Asan Medical Center,

Seoul, South Korea)